Genetic Programming and Evolvable Machines

, Volume 15, Issue 3, pp 281–312

Software mutational robustness

  • Eric Schulte
  • Zachary P. Fry
  • Ethan Fast
  • Westley Weimer
  • Stephanie Forrest
Article

DOI: 10.1007/s10710-013-9195-8

Cite this article as:
Schulte, E., Fry, Z.P., Fast, E. et al. Genet Program Evolvable Mach (2014) 15: 281. doi:10.1007/s10710-013-9195-8

Abstract

Neutral landscapes and mutational robustness are believed to be important enablers of evolvability in biology. We apply these concepts to software, defining mutational robustness to be the fraction of random mutations to program code that leave a program’s behavior unchanged. Test cases are used to measure program behavior and mutation operators are taken from earlier work on genetic programming. Although software is often viewed as brittle, with small changes leading to catastrophic changes in behavior, our results show surprising robustness in the face of random software mutations. The paper describes empirical studies of the mutational robustness of 22 programs, including 14 production software projects, the Siemens benchmarks, and four specially constructed programs. We find that over 30 % of random mutations are neutral with respect to their test suite. The results hold across all classes of programs, for mutations at both the source code and assembly instruction levels, across various programming languages, and bear only a limited relation to test suite coverage. We conclude that mutational robustness is an inherent property of software, and that neutral variants (i.e., those that pass the test suite) often fulfill the program’s original purpose or specification. Based on these results, we conjecture that neutral mutations can be leveraged as a mechanism for generating software diversity. We demonstrate this idea by generating a population of neutral program variants and showing that the variants automatically repair latent bugs. Neutral landscapes also provide a partial explanation for recent results that use evolutionary computation to automatically repair software bugs.

Keywords

Mutational robustness Genetic programming Mutation testing Proactive diversity N-version programming Neutral landscapes 

Copyright information

© Springer Science+Business Media New York 2013

Authors and Affiliations

  • Eric Schulte
    • 1
  • Zachary P. Fry
    • 2
  • Ethan Fast
    • 3
  • Westley Weimer
    • 2
  • Stephanie Forrest
    • 1
    • 4
  1. 1.Department of Computer ScienceUniversity of New MexicoAlbuquerqueUSA
  2. 2.Department of Computer ScienceUniversity of VirginiaCharlottesvilleUSA
  3. 3.Department of Computer ScienceStanford UniversityPalo AltoUSA
  4. 4.Santa Fe InstituteSanta FeUSA

Personalised recommendations